[IEEE 2010 Australasian Telecommunication Networks and Applications Conference (ATNAC 2010) -...

5
Intelligent Features for IMS-based IPTV Nguyen Tai Hung, Nguyen Huu Thanh Department of Communication Engineering Hanoi University of Technology Hanoi, Vietnam {hungtai-fet; thanhnh}@mail.hut.edu.vn AbstractIMS has been widely recognized as the control and signaling framework for delivering of the rich communication and multimedia services to the broadband users. Amongst others, it is deployable as the service (middleware) platform for the interactive and personalized IPTV services. The goal of this paper is to provide an investigation of the advanced use cases in practical perspective. The work has been done at Hanoi University of Technology (HUT) in the scope of its initiatives for NGN research program. Major use- cases, or we called intelligent features, are the advanced electronic service guide, video on demand (VoD), (IPTV) session continuity, and parental control. Development prototypes for each of the use cases are depicted. Keywords - IMS; IMS IPTV; enhanced EPG; Parental Control; Blending Service, Intelligent Features; I. INTRODUCTION IP Multimedia Subsystem (IMS) is the control platform for the next generation network (NGN) architecture currently planned for mobile and fixed multimedia services, standardized by the 3rd Generation Partnership Project (3GPP) [1]. IMS promises a scalable integrated platform that enables new services and provides for the blending of telecommunication and Internet services. Thus, we have chance of using IMS to provide a highly integrated solution for seamless, networked-based media transportation over any end-user device. Since IPTV has been developed and deployed for some time and has gone through numbers of generation with different middleware technologies. When penetrating the comercial market, it is now facing the challenges which demand that IPTV be equipped with more intelligent, personalized and interactive features in a standardized and interoperable way. This paper will explore, in experimental perspective, the possibility of using the IMS framework as the service control plane to provide the interactive and personalized video services to the broadband wired and wireless users. The paper 1 begins with a concise description of the novel IMS based IPTV architecture as well as the setup for an academic IMS test-bed. The main part of the paper will be reserved for presentation of the designated use cases those have been developed in our test-bed environment. Six use cases have been explored and selected for implementation as listed below. 1 This research is conducted under the governmental research project Research and development of service platform for multimedia and value- added services for 3G/4G wireless networks and broadband Internetsupported by NAFOSTED Video on Demand: A standard feature that allows user to request the desired content (a movie, a song or a video clip) any time they wish. Parental Control/Authorization: An intelligent feature that allows the parent to control their children access to (IPTV) broadcast channels and/or on-demand contents by time slot or by identity. Enhanced electronic program guide (EPG): Another feature that provides the (IPTV) viewers with personalized and time-based program guide. Blending services [8]: An feature that allows the composition of telephony/messaging session with video session. II. PREVIOUS WORKS In fact, IMS IPTV is not a new research topic; there are dozen of researches on the issue. For example, on [11] Bruno Chatras et al. investigate how to incorporate IMS components in to IPTV service platforms to provide a flexible framework for novel differentiated services. Traditional telecommunication networks are optimized to have dedicated network functions at certain locations, taking cost, quality of service, reliability and other technical circumstances into account. Load sharing of spread network functions are difficult to realize. With the IP Multimedia Subsystem (IMS), a totally different approach is feasible. It is based on the IP protocol carrying bearer and signaling/control traffic. While bearer traffic functions still need to be optimized with respect to best location, the signaling and control functions of the network could be spread somewhat arbitrarily over the network due to the fact that relatively low traffic between the functions does not significantly impact the optimization result of the total network. This is an additional degree of freedom for assigning such kind of equipment to given locations. It opens up new possibilities for scaling the equipment for total network capacity needs and gives a chance to distribute redundancy geographically, which influences network reliability in a positive way. Moreover, the distributed IMS architecture based on SIP proxy mechanisms provide better means for redundancy and scalability as compared to ”classic soft switches” that still follow a nodal architecture. An important aspect that IMS addresses is its capabilities for blending and orchestration of services; these issues within IMS are addressed in [12]. This is particularly attractive for operators that deploy the so-called blending services, instead of the bundled services, to a large number of young new 2010 Australasian Telecommunication Networks and Applications Conference 978-1-4244-8172-9/10/$26.00 ©2010 IEEE 130

Transcript of [IEEE 2010 Australasian Telecommunication Networks and Applications Conference (ATNAC 2010) -...

Page 1: [IEEE 2010 Australasian Telecommunication Networks and Applications Conference (ATNAC 2010) - Auckland, New Zealand (2010.10.31-2010.11.3)] 2010 Australasian Telecommunication Networks

Intelligent Features for IMS-based IPTV Nguyen Tai Hung, Nguyen Huu Thanh

Department of Communication Engineering

Hanoi University of Technology

Hanoi, Vietnam

{hungtai-fet; thanhnh}@mail.hut.edu.vn

Abstract— IMS has been widely recognized as the control and

signaling framework for delivering of the rich communication and

multimedia services to the broadband users. Amongst others, it is

deployable as the service (middleware) platform for the interactive

and personalized IPTV services. The goal of this paper is to provide

an investigation of the advanced use cases in practical perspective.

The work has been done at Hanoi University of Technology (HUT)

in the scope of its initiatives for NGN research program. Major use-

cases, or we called intelligent features, are the advanced electronic

service guide, video on demand (VoD), (IPTV) session continuity,

and parental control. Development prototypes for each of the use

cases are depicted.

Keywords - IMS; IMS IPTV; enhanced EPG; Parental

Control; Blending Service, Intelligent Features;

I. INTRODUCTION

IP Multimedia Subsystem (IMS) is the control platform for the

next generation network (NGN) architecture currently planned

for mobile and fixed multimedia services, standardized by the

3rd Generation Partnership Project (3GPP) [1]. IMS promises

a scalable integrated platform that enables new services and

provides for the blending of telecommunication and Internet

services. Thus, we have chance of using IMS to provide a

highly integrated solution for seamless, networked-based

media transportation over any end-user device.

Since IPTV has been developed and deployed for some time

and has gone through numbers of generation with different

middleware technologies. When penetrating the comercial

market, it is now facing the challenges which demand that

IPTV be equipped with more intelligent, personalized and

interactive features in a standardized and interoperable way.

This paper will explore, in experimental perspective, the

possibility of using the IMS framework as the service control

plane to provide the interactive and personalized video

services to the broadband wired and wireless users.

The paper1 begins with a concise description of the novel IMS

based IPTV architecture as well as the setup for an academic

IMS test-bed. The main part of the paper will be reserved for

presentation of the designated use cases those have been

developed in our test-bed environment. Six use cases have

been explored and selected for implementation as listed below.

1 This research is conducted under the governmental research project

“Research and development of service platform for multimedia and value-

added services for 3G/4G wireless networks and broadband Internet”

supported by NAFOSTED

• Video on Demand: A standard feature that allows user to

request the desired content (a movie, a song or a video

clip) any time they wish.

• Parental Control/Authorization: An intelligent feature that

allows the parent to control their children access to

(IPTV) broadcast channels and/or on-demand contents by

time slot or by identity.

• Enhanced electronic program guide (EPG): Another

feature that provides the (IPTV) viewers with

personalized and time-based program guide.

• Blending services [8]: An feature that allows the

composition of telephony/messaging session with video

session.

II. PREVIOUS WORKS

In fact, IMS IPTV is not a new research topic; there are dozen

of researches on the issue. For example, on [11] Bruno

Chatras et al. investigate how to incorporate IMS components

in to IPTV service platforms to provide a flexible framework

for novel differentiated services. Traditional

telecommunication networks are optimized to have dedicated

network functions at certain locations, taking cost, quality of

service, reliability and other technical circumstances into

account. Load sharing of spread network functions are

difficult to realize. With the IP Multimedia Subsystem (IMS),

a totally different approach is feasible. It is based on the IP

protocol carrying bearer and signaling/control traffic. While

bearer traffic functions still need to be optimized with respect

to best location, the signaling and control functions of the

network could be spread somewhat arbitrarily over the

network due to the fact that relatively low traffic between the

functions does not significantly impact the optimization result

of the total network. This is an additional degree of freedom

for assigning such kind of equipment to given locations. It

opens up new possibilities for scaling the equipment for total

network capacity needs and gives a chance to distribute

redundancy geographically, which influences network

reliability in a positive way. Moreover, the distributed IMS

architecture based on SIP proxy mechanisms provide better

means for redundancy and scalability as compared to ”classic

soft switches” that still follow a nodal architecture. An

important aspect that IMS addresses is its capabilities for

blending and orchestration of services; these issues within

IMS are addressed in [12]. This is particularly attractive for

operators that deploy the so-called blending services, instead

of the bundled services, to a large number of young new

2010 Australasian Telecommunication Networks and Applications Conference

978-1-4244-8172-9/10/$26.00 ©2010 IEEE 130

Page 2: [IEEE 2010 Australasian Telecommunication Networks and Applications Conference (ATNAC 2010) - Auckland, New Zealand (2010.10.31-2010.11.3)] 2010 Australasian Telecommunication Networks

generation viewers. Being different from the most of the

published papers that does the investigation of IMS IPTV

from architecturing and functional perspectives, our paper

focuses on the exploration of the IMS IPTV new advantages

of highly interaction, personalization and service orchestration

in terms of the viewer’s use cases.

III. IMS-BASED IPTV ARCHITECTURE

A. Novel framework

Figure 1. IMS-based IPTV

Figure 1 shows a high-level functional IPTV network

architecture being supported by an IMS infrastructure. The

model provides multimedia services to the end user by means

of the SIP Application Server (SAS). The SAS implements the

service logics which users interact and request the movies

and/or other online contents in the personalized and intelligent

ways. Important logics have been implemented for the special

features like enhanced electronic program guide (eEPG),

interactive (parental) authorization, blending sessions, etc. The

SAS interacts with the IPTV Terminal Funtionality (IUF) that

handles the display and interactivity functions for viewers.

The IUF also performs functions such as content

encoding/decoding and buffering for both unicast and

multicast streams. The system is divided into a number of

logically separated parts, namely, the home domain, the core

network, the access/transportation network, the service/content

domain and service management.

B. IPTV User Profile

In the IMS IPTV architecture, personalization is an

important feature. To achieve personalization at the

application level (i.e. personalized EPG’s, advertisements, or

even personalized blended communication services), every

user has an IPTV profile. The relation between the IPTV

profile and the IMS profile depends on the availability of a

home IMS gateway. The home gateway is a functional block

with an attached ISIM card reader, which can be deployed in

the residential gateway or any other networked consumer

equipment. The gateway translates home signaling, whether

SIP, UPnP or perhaps pure HTTP to IMS signaling. It also

takes care of NAT traversal and secures connectivity with the

P-CSCF in the IMS domain, as well as identity, device

subscription, and management inside the home domain and

towards the IMS core.

If the household contains a home gateway, the family members can choose whether they want to have full IMS identity, which enables them to have full communication capabilities supported by IMS, or they just opt to have an IPTV profile, that will use the default IMS household identity for authentication purposes. The IPTV profile information that needs to be shared between different IMS services is stored in the IPTV XDMS database. This database is accessed using XCAP, which works over HTTP. These profiles can be shared by different users and other stakeholders within the IPTV system.

C. The Testbed

Figure 2. The HUT Next Generation Test-bed

In the scope of the joint-research project between Hanoi University of Technology and Fraunhofer Institute FOKUS Berlin - Germany, we set up a next generation network test-bed in our lab for the purpose of prototyping new multimedia and rich-feature communication services using IMS framework. The test-bed consists of all three layers: media layer for transportation of media traffic in unicast, multicast and broadcast. The core layer for signaling and session/service control makes use of the FOKUS’ open source IMS Core [3,4,5] that delivers CSCF servers and a light user profile database (HSS). Our project main focus is on the application layer in which we specified and developed prototypes for IP telephony and IP television value added services based on Sailfin platform. A Media Server was also developed at our lab using VLC (VideoLAN) media stack. Besides that we have developed a comprehensive framework and prototype of IMS IPTV Client that is based on the open source IMS Communicator. Finally, several IMS interfaces, namely, Sh, Mw, etc are implemented on our own effort. Figure 2 depicts high level view of our test-bed setup.

131

Page 3: [IEEE 2010 Australasian Telecommunication Networks and Applications Conference (ATNAC 2010) - Auckland, New Zealand (2010.10.31-2010.11.3)] 2010 Australasian Telecommunication Networks

IV. USE CASES

A. Video on Demand

Figure 3. Requesting the on-demand contents

In our research, as depicted on Figure 3, we consider a

scenario in which an IMS user initiates a call to a specific

content (a movie, a song or other resources) locating at the

content provider through an IMS domain. The request can be

routed through different SIP servers (CSCFs); at S-CSCF a

suitable iFC is invoked to forward the request to IPTV AS, the

AS will then proxy the request to MRF (Media Server). Media

Server, after accepting the request, will send back the

successful responses (via AS) as well as the RTP streams

directly to the emulated setop box (STB). In our scenario, we

had implemented the AS based on Sailfin platform; the Media

Server was designed and implemented using VLC

(VideoLAN) RTP stack and oSIP library.

B. Remote Parental Control

A special feature, called Parental Control, had been designed

and implemented which allows the parent to control their child

from requesting and viewing classified contents. In this

scenario, as shown on the Figure 4, the child (Peter) initiates a

call to a specific content (movie and channel), the request will

be forwarded to IPTV AS through IMS Core, the AS, through

the in-house developed Sh interface, will download and check

the service registration of this user and learnt that this user

need the permission (from his parent) in order to view the

requested content. The AS then checks the status of the parent

(Alice) through Presence service and asks for her permission

(via Message method). Depending on the response from Alice,

whether she accepts or denies the request, IPTV AS will either

send back to Peter the denial response (603 Decline) or proxy

the request to Media Server which will subsequently send RTP

stream to Peter’s Client.

C. Enhanced EPG

Personalization is a key feature in the IMS IPTV solution.

In this sense, we have complemented the user profile with a

new XML-formatted [9] service profile for each IMS identity

to contain the personalized information. That leads to another

intelligent feature for IMS-based IPTV, we called Enhanced

EPG. With this feature user will be classified in to different

groups (via subscription) with different service levels.

Figure 4. IPTV Parental Control Feature

In this scenario, users in different classes, after registering

in to the IMS Core, will send Subscribe request to IPTV AS

for the EPG service. The AS then, via Sh interface, will query

the HSS to download the relevant iFC and check the registered

service profile of the requesting user, if the user belongs to the

free service package (no need access control) then AS will

build the relevant channel list and send it, through the payload

(in XML) of 200 OK Response message, back to the

requesting user. In the other hand, if the user belongs to the

group that need access control (e.g. for Premium service

packages) then the AS needs to query the HSS again to obtain

the necessary data to build the suitable channel list and sends

it back to the requesting user. Figure 5 and 6 below show the

basic message exchanges for implementation of these two

scenarios. The access control also provides the time

constrained service profile in which the requesting for

different time slots will receive different channel lists.

Figure 5. Message Flow for EPG of un-controlled access

132

Page 4: [IEEE 2010 Australasian Telecommunication Networks and Applications Conference (ATNAC 2010) - Auckland, New Zealand (2010.10.31-2010.11.3)] 2010 Australasian Telecommunication Networks

Figure 6. Message flow for Enhanced EPG Feature

D. Session Continuity

MN P-CSCF I-CSCF S-CSCF proxy AS

305 use proxy

INVITE(mSCTP)INVITE

INVITE(mSCTP)INVITE(mSCTP)

INVITE(mSCTP)

INVITE(mSCTP)INVITE(mSCTP)

INVITE(mSCTP)100 trying

100 trying100 trying

183 sp183 sp

183 sp183 sp

183 sp

PRACKPRACK

PRACKPRACK

MN PDP

Context Activation 200 OK200 OK

proxy PDP

Context Activation

AS PDP

Context Activation

200 OK200 OK

Visited IMS Domain 1 Home IMS Domain

REFER (AS)

200 OK

MN P-CSCF I-CSCF S-CSCF proxy AS

305 use proxy

INVITE(mSCTP)INVITE

INVITE(mSCTP)INVITE(mSCTP)

INVITE(mSCTP)

INVITE(mSCTP)INVITE(mSCTP)

INVITE(mSCTP)100 trying

100 trying100 trying

183 sp183 sp

183 sp183 sp

183 sp

PRACKPRACK

PRACKPRACK

MN PDP

Context Activation 200 OK200 OK

proxy PDP

Context Activation

AS PDP

Context Activation

200 OK200 OK

Visited IMS Domain 1 Home IMS Domain

REFER (AS)

200 OK

UPDATEUPDATE

180 ringing180 ringing

180 ringing180 ringing

180 ringing

PRACKPRACK

PRACKPRACK

200 OK200 OK

200 OK200 OK

UPDATE

200 OK200 OK

200 OK200 OK

UPDATE

200 OK200 OK

200 OK200 OK

200 OK ACKACK

ACKACK

ACK

mSCTP connection between MN and proxy

TCP/UDPconnection

proxy <> AS

UPDATEUPDATE

180 ringing180 ringing

180 ringing180 ringing

180 ringing

PRACKPRACK

PRACKPRACK

200 OK200 OK

200 OK200 OK

UPDATE

200 OK200 OK

200 OK200 OK

UPDATE

200 OK200 OK

200 OK200 OK

200 OK ACKACK

ACKACK

ACK

mSCTP connection between MN and proxy

TCP/UDPconnection

proxy <> AS

Figure 7. mSCTP connection setup between MN, proxy, AS and other

components in IMS

The issue of session continuity is also studied in our test-bed,

in which we investigated a new approach [10] that allows

handing over of an on-going IPTV session between different

access heterogeneous environments. We propose a new

component in the IMS domain, namely an proxy based on

mSCTP (mobile Stream Control Transmission Protocol) that

acts as an anchor point for soft vertical handover of mobile

nodes, which have multiple physical interfaces (e.g.,

WLAN/UMTS). Principally, our approach facilitates a break-

before-make handover scheme with QoS support, in which the

mSCTP proxy can set up multiple mSCTP tunnels pre-

established via multiple networks. The mSCTP-based proxy

can support QoS provisioning and adaptation for the mobile

nodes when moving in a heterogeneous wireless environment.

The network scenario for session continuity is depicted in

Figure 10.

After registering with a network, let’s say the visited IMS

domain 1 as illustrated in Figure 7, the MN wants to initiate a

call to an IPTV Application Server. At this moment the MN

does not know the existence of an mSCTP-based proxy, thus it

sends an INVITE message with the URI of the service identity

(PSI). Since the MN supports mSCTP in the transport layer, it

would like to set up an mSCTP connection if possible. In our

implementation, instead of sending an INVITE message

informing that the expected transport protocol is TCP or UDP,

mSCTP is specified in the INVITE message on the VIA header

field. Upon receiving the INVITE message, the home S-CSCF

analyses the message and realizes that the MN requests

mSCTP. Instead of forwarding INVITE to the AS, the home S-

CSCF sends a redirect message “305 use proxy” to the MN,

which specifies the URI of the mSCTP-based proxy in the

Contact field. The MN then sends INVITE to the proxy,

requesting to build an mSCTP transport session between the

MN and the mSCTP-based proxy. When the home S-CSCF

process the redirected INVITE message from the MN, it will

also send a REFER to the proxy instructing it to establish a

session from the proxy to the AS.

Following INVITE messages are QoS provisioning steps.

These steps are much like the conventional signaling flow,

except that the proxy should set up two sessions: one between

MN and the proxy with mSCTP tunnel, and the other between

the proxy and the AS. There are three PDP contexts that

should be established for the MN, proxy and AS. The session

setup procedure in our approach is a bit more complex than

usual with 66 messages. In Figure 7, we neglect some

messages that do not play significant role in the signaling

flow. The most advantages of our approach are in the vertical

handover process. The fast handover operation just requires 12

messages in our case versus 46 messages in the conventional

approach. For the handover including QoS setup, it takes 58

messages, 12 messages more when compared to the

conventional approach. If the two mSCTP tunnels already

exist, instead of a lengthy full session establishment, only 2

SIP messages are required to switch the session forth and back

through the two legs of the mSCTP tunnels.

V. SOME EXPERIMENTAL RESULTS

This section highlights some of the various intelligent features

implemented in our test-bed prototype. Figure 8 illustrates a

personalized user portal which provides a different content

meta-data (channel list) to registered user from different

group. Figure 9 shows feature of blending services that allows

displaying the Incoming Call/Message from buddies on the

online TV screen if the watching user is not on the “Do not

133

Page 5: [IEEE 2010 Australasian Telecommunication Networks and Applications Conference (ATNAC 2010) - Auckland, New Zealand (2010.10.31-2010.11.3)] 2010 Australasian Telecommunication Networks

disturb” mode. The viewer then has options of pausing the TV

session to handle the incoming call/message or reject it.

Finally, figure 10 shows our proposed architecture for the

implementation of (IP TV) session continuity.

Figure 8. Channel List for different User Categories of Enhanced EPG

feature

Figure 9. Incoming Call/Message Display

VI. CONCLUSIONS AND FUTURE DIRECTIONS

This paper presents our investigation, in experimental

perspective, of using IMS framework to provide intelligent

features for IPTV services. In particular, it focuses on the video

on demand, remote parental control, blending services, session

handover without interruption and other interactive features

which some of the use cases were discussed and presented

here-above. It shows how SIP [2, 7] signaling and IMS can be

used to provide the Interactive and Blending features for the

entertainment video services.

The initial results promise the great potential of those IMS-based TV interactive and differentiated features, which offer attractive and rich multimedia experiences to the end user. We are currently investigating and developing several other intelligent features of IMS-based TV, namely, the context-based session continuity that allows to seamlessly handover the

IPTV sessions across different screens/terminals on different access networks.

Visited IMS domain 2 - WLANVisited IMS domain 1 - UMTS

Home IMS domain

GGSN PDG

AP

P-CSCF

I-CSCF

HSS

GGSN

S-CSCF

AS

MN

mSCTPProxy

P-CSCF

Data path

Signaling path

RNC

Visited IMS domain 2 - WLANVisited IMS domain 1 - UMTS

Home IMS domain

GGSN PDG

AP

P-CSCF

I-CSCF

HSS

GGSN

S-CSCF

AS

MN

mSCTPProxy

P-CSCF

Data path

Signaling path

RNC

Figure 10. Signaling paths and data paths between MN, mSCTP proxy, AS

and IMS components

VII. ACKNOWLEDGEMENT

The authors would like to acknowledge the National

Foundation for Science and Technology Development

(NAFOSTED) of Vietnam for funding this work.

REFERENCES

[1] TS 23.328, “IP Multimedia Subsystem,” 3GPP, Release 6

[2] J. Rosenberg, et al., “SIP: Session Initiation Protocol,” RFC 3261, IETF, June 2002

[3] F. inc., XDMS: The FOKUS XML Document Management Server,

http://www.fokus.fraunhofer.de/bereichsseiten/.

[4] Open IMS Core Playfround, see http://www.openimscore.org/

[5] SIP Express Router, see http://www.iptel.org/ser

[6] UCT IMS Client, see http://uctimsclient.berlios.de

[7] M. Handley, V. Jacobson, “SDP: Session Description Protocol,” RFC 2327, IETF, April 1998

[8] Service Capability Interaction Manager, see http://en.wikipedia.org/wiki/Service_Capability_Interaction_Manager

[9] J. Rosenberg, The Extensible Markup Language (XML) Configuration Access Protocol, http://tools.ietf.org/html/draft-ietf-simple-xcap-12.

[10] Nguyen Huu Thanh, Le Thi Hang, Vu Van Yem, Ngo Quynh Thu, Nguyen Xuan Dung, “Multimedia Session Continuity with Context-Aware Capability in IMS-based Networks”, in Proceedings of IEEE Sixth International Symposium on Wireless Communication Systems 2009 (ISWCS’09), September 7 – 10, 2009, Siena-Tuscany, Italy

[11] B. Chatras, M. Said. Delivering Quadruple Play with IPTV over IMS. International Conference on Intelligence in Networks (ICIN) 2007, Bordeaux, France

[12] Beck, Andre , Bob Ensor, Jairo Esteban, 2007, "IMS and IPTV Service Blending -Lessons and Opportunities," Journal of the Institute of Telecommunications Professionals

134